ggplot2 Resources

Barplots

ggplot(data=incex) + geom_bar(mapping = aes(x = occ))

  • geom_bar() uses stat_count() by default. We can reproduce the barplot using stat_count().
ggplot(incex) + stat_count(aes(x = occ))

  • If you have precomputed data, use identity stat.
occ_counted <- incex %>% count(occ, name = 'count') 
occ_counted
   occ count
1  low   595
2 med.   700
3 high   627
ggplot(occ_counted) + geom_bar(aes(x = occ, y = count), stat = 'identity')

ggplot(occ_counted) + geom_col(aes(x = occ, y = count)) # shortcut

  • Use after_stat() to modify mapping from stats (here we calculate %).
ggplot(incex) + geom_bar(aes(x = occ,
  y = after_stat(100 * count / sum(count))))

  • A stacked plot with frequencies
ggplot(occ_counted) + geom_bar(aes(x="", y = count, fill=occ), stat = "identity")

  • Based on the vector type of the variable sex, a discrete colour scale is picked.
ggplot(incex) + geom_bar(aes(x = occ, fill=sex))

  • You may want to change colors.
ggplot(incex) + geom_bar(aes(x = occ, fill=sex)) + 
  scale_fill_brewer(palette="Dark2")

  • You may want to choose colors manually.
ggplot(incex) + geom_bar(aes(x = occ, fill=sex)) + 
  scale_fill_manual(values=c("#999999", "#E69F00"))

  • A stacked plot with proportions. position = "fill" makes “proportion bars”.
ggplot(occ_counted) + geom_bar(aes(x="", y = count, fill=occ), stat = "identity",   position="fill") + labs(y="proportion")

ggplot(incex) + geom_bar(aes(x = occ, fill=sex), position = "fill") + 
  labs(x="occupational status", y="proportion", title="stacked plot")

Pie Chart

  • A polar coordinate system interprets x and y as radius and angle.
occ_counted <- incex %>% count(occ, name = 'count') 
p <- ggplot(occ_counted) + geom_bar(aes(x="", y = count, fill=occ), 
                                    stat = 'identity', width=1)
p + coord_polar(theta = "y")

  • Let’s change theme. Here, theme_void() removes background, grid, numeric labels.
p + coord_polar(theta = "y") +
  theme_void()

Histogram

  • Draw a histogram with a continuous variable.
ggplot(incex) + geom_histogram(aes(x = income), bins = 20) + theme_minimal()

  • Density Plot
ggplot(incex) + geom_density(aes(x = income)) + theme_minimal()

  • Draw the histogram and density plots together.
ggplot(incex) + geom_histogram(aes(x = income, y=..density..), bins=20) + 
  geom_density(aes(x = income), col="red") + theme_minimal()

  • We can globally define mapping parameters inside ggplot.
ggplot(incex, aes(x = income, y=..density..)) + 
  geom_histogram(bins=20) + geom_density(col="red") + theme_minimal()

  • Histogram with groups
ggplot(incex) + geom_histogram(aes(x = income, fill=sex), 
  bins=20, alpha=0.7) + theme_bw()

Boxplot

  • Draw a boxplot.
ggplot(incex, aes(x = income)) + geom_boxplot() + theme_bw()

  • Remove everything on the y-axis including the texts and tick marks.
ggplot(incex, aes(x = income)) + geom_boxplot() + theme_bw() +
      theme(axis.text.y=element_blank(),
            axis.ticks.y=element_blank())

  • Draw a boxplot by the variable sex.
ggplot(incex, aes(x = income, y = sex)) + geom_boxplot() + theme_bw()

Violin Plot

  • A violin plot displays distributions of numeric data for one or more groups using density curves. The width of each curve corresponds with the approximate frequency of data points in each region.
ggplot(incex, aes(x = income, y = sex)) + geom_violin() + theme_bw()

  • You may want to overlay data points.
ggplot(incex, aes(x = income, y = sex)) + geom_violin() + 
  geom_point(alpha=0.2) + theme_bw()

  • You may want to jitter data points.
ggplot(incex, aes(x = income, y = sex)) + geom_violin() + 
  geom_jitter(width = 0.1, height = 0.1, alpha=0.2) + theme_bw()

Quantile-Comparison Plot

  • Compare the observed (empirical) sample distribution of a variable with a theoretical distribution.
ggplot(incex, aes(sample = income)) + stat_qq() + stat_qq_line() + theme_classic()

Scatterplot

  • geom_point() with two continuous variables
ggplot(incex, aes(x = age, y = income)) + geom_point() + theme_bw()

  • Jittered scatterplot; “jittering” adds a small random quantity (uniformly distributed) to each observation.
ggplot(incex, aes(x = age, y = income)) + geom_jitter() + theme_bw()

  • Scatterplot with different transparency levels; alpha controls the degree of transparency for data points.
ggplot(incex, aes(x = age, y = income)) + geom_point(alpha=0.3) + theme_bw()

  • Scatterplot with groups; Use the argument col or shape inside aes().
ggplot(incex, aes(x = age, y = income, col=sex)) + 
  geom_point(alpha=0.3) + theme_bw()

ggplot(incex, aes(x = age, y = income, shape=sex)) + 
  geom_point(alpha=0.3) + theme_bw()

Facets

  • Two facet functions for splitting data by categories

  • facet_wrap() : “wraps” a 1d ribbon of panels into 2d.

  • facet_grid() : produces a 2d grid of panels defined by variables which form the rows and columns.

facet_wrap()

  • Use facet_wrap() with one categorical variable
ggplot(incex, aes(x = age, y = income)) + geom_point() + 
  facet_wrap(~ sex)

  • Use facet_wrap() with two categorical variables: sex and edu
ggplot(incex, aes(x = age, y = income)) + geom_point() + 
  facet_wrap(sex ~ edu)

  • Use facet_wrap() with two categorical variables: occ and edu
ggplot(incex, aes(x = age, y = income)) + geom_point() + 
  facet_wrap(occ ~ edu)

facet_grid()

  • Use facet_grid() with variables sex and edu.
ggplot(incex, aes(x = age, y = income)) + geom_point() + 
  facet_grid(sex ~ edu)

  • Use facet_grid() with variables occ and edu.
ggplot(incex, aes(x = age, y = income)) + geom_point() + 
  facet_grid(occ ~ edu)

Themes

  • Try different arguments for theme!
ggplot(incex) + geom_bar(aes(y = occ)) + facet_wrap(~ sex) +
    labs(title = "Number of occupational status by gender",
         caption = "source: DS3003, Fall 2021",
         x = NULL,
         y = NULL) +
    theme_minimal() +
    theme(
      strip.text = element_text(face = 'bold', hjust = 0),
      plot.caption = element_text(face = 'italic'),
      panel.grid.major = element_line('white', size = 0.5),
      panel.grid.minor = element_blank(),
      panel.grid.major.y = element_blank(),
      panel.ontop = TRUE
    )

Beyond ggplot2

  • ggplot2 is huge! About 50 geoms, 25 stats, 60 scales.

  • Many extensions are very niche specific and developed by experts in the field.

  • 100 registered extensions available to explore (https://exts.ggplot2.tidyverse.org)

  • e.g., for plot compositions, you might want to use gridExtra::grid.arrange(), ggpubr::ggarrange(), cowplot::plot_grid(), or patchwork.

Example: package gridExtra

library(gridExtra)
p1 <- ggplot(incex) + geom_bar(aes(x = occ, fill=sex)) # barplot
p2 <- ggplot(incex) + geom_histogram(aes(x=income), bins=10) # histogram
p3 <- ggplot(incex) + geom_point(aes(x = age, y=income)) # scatterplot

grid.arrange(p1, p2, p3, nrow=1)

Example: package patchwork

library(patchwork)
p1 + p2 + p3

( p1 + p2 ) / p3

Example: package GGally

library(GGally)
ggpairs(incex[, c('income', 'oexp', 'age', 'edu')])